Diffusion Tensor Imaging of Cerebral White M atter Technique, Anatomy, and Pathologic Patterns Asim F. Choudhri, MDa,b,c,d,*, Eric M. Chin, BSa, Ari M. Blitz, MDe, Dheeraj Gandhi, MDf,g,h KEYWORDS  Diffusion tensor imaging  White matter  Physiologic imaging  Magnetic resonance imaging  Anatomy  Brain

KEY POINTS  Diffusion tensor imaging is a magnetic resonance (MR) imaging technique that allows visualization of location, orientation, and integrity of white matter pathways.  Mathematical constructs underlying diffusion tensor imaging are complex, but a basic understanding can guide interpretation.  Interpretation of diffusion tensor parameter maps in conjunction with conventional MR imaging techniques can aid in diagnosis of white matter development and disorders.  Processing diffusion tensor imaging data via diffusion tensor fiber tracking tractography allows mapping of individual tracts, which can be useful in surgical planning.

The use of diffusion-weighted imaging (DWI) is well established in the rapid diagnosis and evaluation of cerebral infarction,1 as well as for identifying lesions such as epidermoids and, more recently, characterizing the cellularity of tumors.2 Diffusion tensor imaging (DTI), an advanced form of DWI, is an important tool in evaluating white matter anatomy and in pathology.3,4 Although originally a research tool and only used in academic centers,

DTI has become a valuable part of the clinical evaluation of brain development, and in surgical planning for brain tumors.5–8 The use of DTI has recently been investigated in the spinal cord.9–11 This article reviews the techniques of DTI and provides a practical approach for clinical implementation and interpretation. The anatomy of the white matter tracts, previously largely unseen by the radiologist, is increasing in importance (Table 1). This article reviews the fundamental

Disclosures: None. a Department of Radiology, University of Tennessee Health Science Center, 848 Adams Avenue, G216, Memphis, TN 38103, USA; b Department of Neurosurgery, University of Tennessee Health Science Center, 848 Adams Avenue, G216, Memphis, TN 38103, USA; c Department of Ophthalmology, University of Tennessee Health Science Center, 848 Adams Avenue, G216, Memphis, TN 38103, USA; d Le Bonheur Neuroscience Institute, Le Bonheur Children’s Hospital, 848 Adams Avenue, G216, Memphis, TN 38103, USA; e Division of Neuroradiology, Department of Radiology and Radiological Science, Johns Hopkins University, 600 N Wolfe Street, B100, Baltimore, MD 21287, USA; f Division of Neuroradiology, Department of Radiology, University of Maryland, 22 S Greene Street, Baltimore, MD 21201, USA; g Department of Neurology, University of Maryland, 22 S Greene Street, Baltimore, MD 21201, USA; h Department of Neurosurgery, University of Maryland, 22 S Greene Street, Baltimore, MD 21201, USA * Corresponding author. Department of Radiology, Le Bonheur Children’s Hospital, 848 Adams Avenue, G216, Memphis, TN 38103. E-mail address: [email protected] Radiol Clin N Am 52 (2014) 413–425 http://dx.doi.org/10.1016/j.rcl.2013.11.005 0033-8389/14/$ – see front matter Ó 2014 Elsevier Inc. All rights reserved.

radiologic.theclinics.com

INTRODUCTION

414

Choudhri et al

Table 1 Major white matter tracts Tract

Origin

Destination

Course

Function

Corticospinal tract

Precentral gyrus

Anterior horn cells of spinal cord

Motor control for body

Corticobulbar tract

Precentral gyrus (inferiorly)

Pons

Spinothalamic tract

Posterior horn cells of spinal cord

Rostral ventromedial thalamus

Geniculocalcarine tract

Lateral geniculate nucleus

Juxtacalcarine occipital cortex

Forceps major

Occipital cortex

Forceps minor

Frontal pole cortex

Contralateral occipital cortex (homotopic) Contralateral occipital cortex (homotopic)

Traverses PLIC, cerebral peduncle, pyramidal decussation, lateral columns of spinal cord Traverses PLIC, cerebral peduncle, pyramidal decussation Lateral and anterior tracts in cord, anterior white commissure decussation, posterolateral pons, and midbrain Lateral margin of atrium and occipital horn of lateral ventricles Splenium of corpus callosum Genu of corpus callosum

Frontal association

Central Tegmental Tract Ascending fibers

Solitary tract nucleus

Midbrain and pons

Ascending taste fibers

Ventral posteromedial nucleus of thalamus

Motor control for cranial nerves Sensory information from body to the thalamus

Visual fibers

Visual association

Descending fibers (rubroolivary tract) Transverse pontocerebellar fibers

Red nucleus (parvocellular)

Inferior olivary nucleus

—

Cerebellum

Contralateral cerebellum

Basilar pons and middle cerebellar peduncle

Arcuate fasciculus

Wernicke area (typically in region of planum temporale) Multidirectional fiber bundles connecting frontal, parietal, occipital, and posterior temporal lobes Bidirectional between temporal and occipital lobes Limbic system (including amygdala and hippocampus)

Broca area (typically in region of pars opercularis and pars triangularis) Multidirectional fiber bundles connecting frontal, parietal, occipital, and posterior temporal lobes Bidirectional between temporal and occipital lobes Orbitofrontal cortex

Frontoparietotemporal opercular

Superior longitudinal fasciculus

Inferior longitudinal fasciculus Uncinate fasciculus

Frontal, temporal, and parietal lobes, lateral to corona radiata

Temporal and occipital lobes, lateral to lateral ventricle Temporal uncus and stem, inferior frontal lobes

Connection to contralateral cerebellar hemisphere Not fully understood; impairment likely involved in MSA-C32 Conveys information about the sound within words (not the meaning) Intrahemispheric association bundles

Intrahemispheric association bundles Likely extension of limbic system

Abbreviations: MSA-C, multisystem atrophy with predominant cerebellar signs; PLIC, posterior limb of the internal capsule.

DTI of Cerebral White Matter 415

416

Choudhri et al anatomic structures visible on DTI that are important to the understanding of white matter pathways and associated pathologic conditions (Table 2).

axons) thus demonstrate highly anisotropic diffusion characteristics.

DTI: Theory and Techniques

IMAGING TECHNIQUE Techniques of DTI Although the physics of DTI can seem daunting, a conceptual understanding is facilitated by understanding the terminology. Perhaps the most important term to understand is anisotropy, a word that is less complex than it might seem. “Iso” is the prefix meaning same and “tropic” is a suffix meaning turning toward or changing, and thus isotropic means moving the same in all directions. In the absence of barriers to movement, a molecule of water randomly moves in all directions in a process known as brownian motion, giving the system isotropic diffusion characteristics. The prefix “an” means not, and therefore anisotropic means movement that is not the same in all directions. Anatomic structures, such as white matter tracts, in which diffusion is highly constrained in some directions (eg, perpendicular to the direction of the contained axons) but remains unrestricted in other directions (eg, parallel to the direction of the

DTI is a magnetic resonance (MR) imaging technique that allows quantification in tissue of water diffusion characteristics such as anisotropy and the direction of greatest diffusivity. In white matter imaging, these parameters can be used to infer the paths and integrity of individual white matter tracts.1,3,12 In white matter, the high degree of anisotropy observed13 is thought to be caused by bundled parallel axons. Perpendicular to the long axis of the axons (ie, radial movement), diffusion is impeded by the axonal cell membrane and myelin sheath. In contrast, along the long axis of the axons (ie, axial movement), diffusion is less constrained. Thus, if axonal microstructure or myelination is compromised by a disease process, a corresponding decrease in anisotropy can be measured. In DWI, a pulsed field gradient is used (in contrast with the normal homogeneous magnetic field used for T1-weighted and T2-weighted

Table 2 Major white matter anatomic structures White Matter Structure Corpus callosum Genu/rostrum Body Isthmus Splenium Internal capsule ALIC PLIC

Optic nerve and tract Cerebral peduncles

Cerebellar peduncles Superior

Middle Inferior Anterior commissure

Fiber Tracts Contained Within

Function

Forceps minor — — Forceps major

Interhemispheric Association  Frontal  Frontal/perirolandic  Perirolandic/precuneus  Occipital

Various Corticospinal tract Corticobulbar tract Optic tract Optic pathway Corticospinal tract Corticobulbar tract Spinothalamic tracts

Various Motor for body Motor for head/face Visual information Visual information Motor to body Motor to head/face Sensory fibers

Ventral spinocerebellar tract, dentatothalamic tract, trigeminocerebellar fibers Pontocerebellar fibers Dorsal spinocerebellar tract, olivocerebellar tract —

Proprioception, various

Various Various Commissural fibers connecting temporal lobes and amygdalae

DTI of Cerebral White Matter images); this causes water molecules to gyrate throughout the tissue. In addition, all diffusion sequences apply a spin echo by producing another gradient pulse in the opposite direction to refocus the water molecules (ie, bringing them back into phase). Because diffusion has a defocusing effect, increased diffusivity nonlinearly attenuates the signal received. However, because diffusionweighted sequences rely on a field gradient, they are sensitive to diffusion only in a single direction. Sequences used in DTI are an extension of those used in DWI. By repeating the sequence 6 or more times while varying the gradient direction, what is known as the diffusion tensor can be reconstructed: a mathematical entity describing the diffusion characteristics at each voxel imaged, roughly corresponding with the summation of all vectors of water molecule diffusion in the voxel. This tensor is used to calculate the various metrics described later (eg, fractional anisotropy [FA]; axial, radial, and mean diffusivity). In general, current clinical use of DTI can be classified as (1) direct semiquantitative interpretation of maps of these metrics to assess axonal health and myelin status, (2) use of these metrics as an adjunct with other modalities (such as conventional T1-weighted and T2-weighted sequences) to better characterize tissue abnormalities, and (3) use of tractography (described later) to identify the paths of key white matter tracts in surgical planning. A minimum of 6 directions of encoding are required to obtain directional information and FA calculation, but a larger number of encoding directions improves white matter characterization. High-resolution clinical studies, such as for tumor surgical planning, often use 15 to 30 directions of encoding, and research studies at times use more than 100 directions of encoding.

FA FA is a dimensionless value that describes the extent of unidirectional water movement, and is scaled from 0 to 1. An FA of 0 indicates no anisotropy in water movement, which is the case for true isotropic movement such as is expected in a glass of water, or within a collection of cerebrospinal fluid such as the ventricular system. An FA of 1 indicates purely unidirectional movement of water, something that does not typically exist in organisms. However, parallel white matter fibers without significant interstitial space, such as in the myelinated corpus callosum, can have an FA of 0.85 or higher (Fig. 1A). FA increases as the brain myelinates.14,15 Lower FA values do not necessarily indicate decreased white matter integrity, but only less unidirectional water motion. As an example, crossing

white matter fibers in the deep white matter may be fully myelinated and intact; however, because of the multidirectional orientation, FA is less than if the same fibers were parallel. FA is also decreased if a greater percentage of a voxel is interstitial space, where there is more isotropic movement and a corresponding lower FA. Evaluating FA on a quantitative basis can be done comparing homotopic structures between the two hemispheres. FA values are typically symmetric, and any asymmetry should raise suspicion for an abnormality, typically with the abnormal side showing a lower FA value. FA can be computed on a structure-bystructure basis, such as for the posterior limb of the internal capsule (PLIC) and compared with age-related normative values. A histogram analysis can be performed of FA on either a wholebrain or a regional basis. Although the region is typically an anatomically defined area (voxelbased analysis), such as the PLIC, histogram analysis can also be performed on a white matter bundle that does not have defined anatomic boundaries (eg, the arcuate fasciculus) in a process known as tract-based spatial statistics (TBSS). Identifying the tract for TBSS is typically performed by diffusion tensor fiber tracking (DT-FT), which is discussed further later. More complex (but automated) FA analyses are being investigated to quantify progression of myelination.16 Display of FA maps can take place in 3 ways. First is a map on which the grayscale value represents the FA value (see Fig. 1A); to facilitate windowing and leveling on picture archiving and communications systems (PACS), FA maps are typically scaled from 0 to 1000, or 0 to 10,000. Thus, a value in the corpus callosum of 8200 likely indicates an FA of 0.82; confirmation of this should be made for MR and PACS vendor combinations before clinically using these values. Color encoding can be performed based on a lookup table (LUT), in which different ranges of FA values are given different colors (see Fig. 1B), and that can be created from any grayscale FA DICOM dataset; however, typically these are not used clinically. Possibly the most important visual display mechanism of FA data is a directionally encoded (DE) FA map (DE-FA; see Fig. 1C), commonly referred to as color-coded FA maps but described here as DE-FA to distinguish them from the LUT-based color FA maps. DE-FA maps cannot be created solely from a grayscale FA map without additional information (derived from the source diffusion images) about the predominant direction of water movement in a voxel. In DE-FA, the FA value of a voxel is represented by brightness, so voxels

417

418

Choudhri et al

Fig. 1. (A) FA, (B) color FA, (C) directionally encoded (DE) FA, (D) mean diffusivity, (E) axial diffusivity, (F) radial diffusivity.

with high FA appear brighter. The color used for each voxel is based on the direction of the primary axis of water diffusion, with the convention stating that red represents movement along the transverse axis, green along the anterior-posterior axis, and blue along the craniocaudal axis. The features on a scan that most clearly show the directional color scheme are the red commissural fibers of the corpus callosum, the green fibers of the optic radiations (or geniculocalcarine tract), and the blue appearance of the descending fibers of the corticospinal tract within the PLIC.

In DWI, the rate of signal decay is calculated between 2 different time values, typically 0 and 1000 milliseconds (ie, B-0 and B-1000). A similar value exists for DTI, which takes the mean value for the diffusion coefficient among the various encoding directions, a value known as mean diffusivity (MD; see Fig. 1D). MD can be helpful in clinical evaluations in which DTI is performed instead of DWI, and it is required to determine whether bright signal on B-1000 images is related to reduced water diffusion (low apparent diffusion coefficient [ADC] and MD values) or to increased water

DTI of Cerebral White Matter content (T2 shine-through, which has high ADC and MD values). However, DTI offers additional information: diffusion coefficients can be calculated for the rate of water movement along the primary axis of water movement, known as axial diffusivity

(AxD; see Fig. 1E). The rate of diffusion in a plane perpendicular to the long axis can be calculated, representing a value known as radial diffusivity (RD; see Fig. 1F). Note that the corpus callosum has free diffusion of water along the long axis

Fig. 2. (A) DE-FA of bilateral occipital cortex with a seed location placed in the right occipital deep white matter. (B) Deterministic DT-FT of the forceps major (red line) beginning from the seed location and progressing toward the splenium of the corpus callosum. (C) Deterministic DT-FT of the forceps major (red line) extending to a homotopic location in the left occipital lobe, as well as the geniculocalcarine tract (pink line) extending toward the lateral geniculate nucleus. (D) Deterministic DT-FT, similar to parts A to C, but with a target region of interest (ROI) placed in the left occipital lobe (blue circle). The forceps major (red line) is unchanged from parts A to C, but the geniculocalcarine tract is no longer present. (E) Axial T1-weighted (T1W) image with DE-FA overlay, as well as overlay of DT-FT mapping the forceps major. The DT-FT is calculated by deterministic tractographic analysis with the ROI placed in the right occipital white matter showing connection to the contralateral occipital lobe through the splenium of the corpus callosum.

419

420

Choudhri et al and reduced water movement perpendicular to the long axis. The significant difference between these two values is related to the same properties that result in a high FA. If AxD and RD are equal, then FA 5 0. Gray matter has intermediate but similar AxD and RD, and a corresponding low

FA. AxD and RD are important concepts, but are not routinely used in clinical practice. Although DTI parameters such as DE-FA are related to the direction and integrity of white matter tracts, perhaps the most intriguing uses of DTI are related to DT-FT. DT-FT, a technique

Fig. 3. Axial DE-FA maps at multiple levels (superior to inferior) showing various tracts, including the corticospinal tract (CST; A–I), corpus callosum (CC; B, C); forceps major (FMaj; C, D), forceps minor (FMin; C, D), PLIC (C, D), cerebral peduncles (CP; E, F), superior cerebellar peduncle (SCP; G), decussation of the superior cerebellar peduncle (SCP-D; G), middle cerebellar peduncle (MCP; H, I), inferior cerebellar peduncle (ICP; I), transverse pontocerebellar fibers (TPF; H, I), central tegmental tract (CTT; H, I), geniculocalcarine tract (GCT; D, E), uncinate fasciculus (UF; F, G), superior longitudinal fasciculus (SLF; B), and inferior longitudinal fasciculus (ILF; D–F).

DTI of Cerebral White Matter commonly referred to as tractography, is a process whereby the integrity and directions of white matter in adjacent voxels can be processed to postulate the location of white matter fiber pathways. To do this, a starting point is chosen (Fig. 2A), known as a DT-FT seed. Knowing that the voxel anterior to this has a predominantly anterior-posterior motion (indicated by green), a line can be drawn suggesting a possible white matter tract taking this course (see Fig. 2B); in this case the fibers of the forceps major. A similar set of connections can be made for the geniculocalcarine tract (ie, optic radiations; pink line in Fig. 2C). By selecting a target region of interest (ROI) in the left occipital lobe, only fibers between the seed and target are included (see Fig. 2D). DT-FT often shows multiple fibers (see Fig. 2E). Once tracts are identified, they can be overlaid on structural imaging for purposes such as surgical planning, or can be used to define an ROI for quantitative analysis of FA data in TBSS. To know which voxels to connect and which to not connect, 2 parameters must be determined. First is an FA threshold. Whichever threshold is chosen, if a voxel has an FA value less than this it is assumed that the proposed tract does not continue through that voxel. The second parameter is the

Fig. 3. (continued)

angle threshold. White matter tracts do not necessarily travel in a straight line, but they do not typically make sharp turns either. A determination can be made for an angle threshold less than which it is assumed that the fiber tract is continuous, and more than which it is assumed that a white matter tract does not continue. This method of DT-FT is known as deterministic DT-FT, and is the most commonly used in clinical settings. Research protocols may use probability theory to determine the most likely course of white matter tracts; a method known as probabilistic tracking, which is intended to have improved performance in following crossing fibers. Common FA thresholds for deterministic tracking are approximately 0.2 to 0.3, and common angle thresholds are approximately 20 to 40 . Changing these parameters can change the outcome of DT-FT significantly, which, coupled with variability in seed and target locations and DTI acquisition parameters, can potentially change results and clinical outcomes.

ANATOMY When evaluating conventional MR, white matter is often evaluated based on structural features, such

421

422

Choudhri et al as the corpus callosum and the internal capsule. DTI and DT-FT allow further evaluation of specific tracts within these structures that may perform different functions. For instance, the PLIC carries motor innervation to the cranial nerves through the corticobulbar tract; motor innervation to the body through the corticospinal tract; and, in the posterior-most portion, carries visual fibers within the geniculocalcarine tract. Innumerable white matter pathways exist within the cerebral hemispheres; however, they can largely be placed into 3 categories. There are commissural fibers, which traverse the midline to the contralateral hemisphere; association fibers, which communicate within a hemisphere; and projection fibers, which leave the cerebral hemispheres and communicate with the brainstem, cerebellum, and spinal cord. Corona radiata and centrum semiovale are terms that are commonly misused. The corona radiata represents the white matter that extends from the internal capsule superiorly into the cerebral

hemispheres. Centrum semiovale represents all the white matter in a cerebral hemisphere. Therefore corona radiata is a subset of centrum semiovale, and the centrum semiovale also contains commissural and intrahemispheric association fibers. The distinction between these two entities is unrelated to their relationship to the lateral ventricles, contrary to commonly misstated definitions. Further confusion about white matter anatomy comes from the sometimes alternating use of descriptions based on its structure, function, and specific tracts. For instance, the optic radiations are structures that perform the function of the visual pathway, and the specific tract is known as the geniculocalcarine tract because the fibers connect the lateral geniculate nucleus to the juxtacalcarine occipital cortex. Major white matter structures and tracts can be identified on DE-FA maps (Fig. 3). A comprehensive review of white matter anatomy is beyond the scope of this article, but detailed anatomic references are available.17–20

Fig. 4. (A) Axial fluid-attenuated inversion recovery (FLAIR) image of a hyperintense lesion in the inferior aspect of the left precentral gyrus. (B) Coronal T1W image with DE-FA overlay showing diminished FA within infiltrated white matter, without signs of displacement of white matter tracts, suggesting an infiltrative tumor. (C) Coronal T1W image with DT-FT overlay showing diminished extension of commissural fibers to the involved parenchyma, identified by seed-ROI placement in the midbody of the corpus callosum.

DTI of Cerebral White Matter IMAGING FINDINGS/PATHOLOGIC CONDITIONS A common use of DTI and DT-FT is for presurgical planning in patients with brain lesions presumed to be tumors.7 Some tumors have circumscribed margins and some have indistinct infiltrating margins. DTI may be able to determine whether a tumor that distorts structural anatomy infiltrates white matter tracts or only displaces them (Fig. 4). The concept is that, if there is infiltration of a white matter tract, resection will disrupt whichever function is performed by the infiltrated tract(s), whereas the function of displaced tracts will likely be spared. The identification of tracts potentially infiltrated by neoplasm allows the neurosurgeon to have an informed discussion with the patient regarding whether or not they will be resected, allowing some degree of prediction of the associated potential postoperative neurologic deficits. At times, the surgeon and patient may elect to perform a full resection, with understanding of the possible rehabilitation required because of the disrupted pathway. At other times, surgeons may selectively spare portions of the tumor, anticipating treating those portions with adjunctive measures such as chemotherapy and/ or radiation, or possibly just close observation. DT-FT can be combined with other physiologic imaging techniques such as blood oxygen level– dependent (BOLD) functional MR (fMR) imaging.

fMR imaging may identify an area of cortex that controls a motor or language task (Fig. 5). Although DT-FT is most commonly used to localize white matter fibers or confirm their integrity, it can also be used to confirm disconnection, such as after a corpus callosotomy (Fig. 6).21–23 Vasogenic edema decreases FA by increasing the water content in the interstitial space. There is facilitated diffusion in all directions, in contrast with cytotoxic edema, which has reduced water diffusion in all directions. Decreased FA can also be seen in the setting of wallerian degeneration.24 FA changes as the pediatric brain matures, increasing as the brain myelinates.13,14 An altered FA maturation pattern has been shown in multiple white matter pathways in patients with autism spectrum disorder (ASD),6 with accelerated maturation and FA greater than controls at 6 months of age followed by an early plateau with FA lower than controls by 24 months of age. Although these trends have been shown on a population basis, threshold values have not yet been validated for this to be used as a tool for early diagnosis of ASD. In patients with tuberous sclerosis complex (TSC), there is decreased FA in the white matter of cortical tubers and their associated white matter migration lines.5,25 Studies have shown that treatment with mammalian target of rapamycin (mTOR) inhibitors such as everolimus can partially reverse these FA abnormalities.26 White matter

Fig. 5. (A) Axial FLAIR image with BOLD-fMR imaging overlay of areas of cortical activation corresponding with expressive language (Broca area, in pink) in the left pars opercularis and pars triangularis. DT-FT performed using the language centers as seed locations identifies the arcuate fasciculus. (B) Axial FLAIR image showing DT-FT of the arcuate fasciculus coursing beneath a left inferior perirolandic mass. (C) Sagittal T1W image showing DT-FT of the arcuate fasciculus coursing posteriorly to the region of the planum temporale. (D) Lateral projection of a three-dimensional rendering of the brain with cutout of the lateral aspect of the left hemisphere, with DT-FT of the arcuate fasciculus, showing a thicker bundle of fibers than is typically seen on any two-dimensional planar view.

423

424

Choudhri et al Fig. 6. (A) Sagittal T1W image in a patient after a splenium-sparing corpus callosotomy shows non-visualization of nearly all of the corpus callosum, and a normal appearance of the splenium. (B) Axial DE-FA map shows an absence of transversely oriented red diffusion within the genu of the corpus callosum, but preservation of the splenium. (C) Posterior projection DT-FT with ROI placed in the right frontoparietal deep white matter shows no commissural fibers traversing the midline through the genu, body, or isthmus of the corpus callosum. (D) Superior projection DT-FT performed with ROI in the right occipital deep white matter shows preservation of fibers of the forceps major traversing the midline within the splenium of the corpus callosum.

abnormalities such as this may be related to altered functional connectivity and increased incidence of ASD in patients with TSC.5,27,28 Atrophic changes in the transverse pontocerebellar fibers are related to the hot-cross-buns sign seen in multiple system atrophy, which can be shown on DTI.29–31

SUMMARY DTI is a powerful tool for evaluating white matter anatomy and associated abnormalities. DTI was previously a research tool, but is entering clinical practice. Understanding the terminology and basic concepts of this technique can aid in the clinical implementation and interpretation of this blend of structural and physiologic white matter evaluation.

REFERENCES 1. Le Bihan D, Breton E, Lallemand D, et al. MR imaging of intravoxel incoherent motions: application to diffusion and perfusion in neurologic disorders. Radiology 1986;161:401–7. 2. Rumboldt Z, Camacho DL, Lake D, et al. Apparent diffusion coefficients for differentiation of cerebellar tumors in children. AJNR Am J Neuroradiol 2006; 27(6):1362–9. 3. Pierpaoli C, Jezzard P, Basser PJ, et al. Diffusion tensor MR imaging of the human brain. Radiology 1996;201:637–48.

4. Lee SK, Kim DI, Kim J, et al. Diffusion-tensor MR imaging and fiber tractography: a new method of describing aberrant fiber connections in developmental CNS anomalies. Radiographics 2005;25: 53–65 [discussion: 66–8]. 5. Krishnan ML, Commowick O, Jeste SS, et al. Diffusion features of white matter in tuberous sclerosis with tractography. Pediatr Neurol 2010;42:101–6. 6. Wolff JJ, Gu H, Gerig G, et al. Differences in white matter fiber tract development present from 6 to 24 months in infants with autism. Am J Psychiatry 2012;169(6):589–600. 7. Pillai JJ, Zaca D, Choudhri A. Clinical impact of integrated physiologic brain tumor imaging. Technol Cancer Res Treat 2010;9:359–80. 8. Choudhri AF, Narayana S, Rezaie R, et al. Same day tri-modality functional brain mapping prior to resection of lesion involving eloquent cortex. Neuroradiol J 2013;26(5):548–54. 9. Thurnher MM, Law M. Diffusion-weighted imaging, diffusion-tensor imaging, and fiber tractography of the spinal cord. Magn Reson Imaging Clin N Am 2009;17:225–44. 10. Thurnher MM, Bammer R. Diffusion-weighted magnetic resonance imaging of the spine and spinal cord. Semin Roentgenol 2006;41:294–311. 11. Ducreux D, Lepeintre JF, Fillard P, et al. MR diffusion tensor imaging and fiber tracking in 5 spinal cord astrocytomas. AJNR Am J Neuroradiol 2006;27: 214–6. 12. Mukherjee P, Berman JI, Chung SW, et al. Diffusion tensor MR imaging and fiber tractography: theoretic

DTI of Cerebral White Matter

13.

14.

15.

16.

17.

18.

19. 20. 21.

22.

underpinnings. AJNR Am J Neuroradiol 2008;29: 632–41. Moseley ME, Cohen Y, Kucharczyk J, et al. Diffusionweighted MR imaging of anisotropic water diffusion in cat central nervous system. Radiology 1990;176: 439–45. Provenzale JM, Isaacson J, Chen S, et al. Correlation of apparent diffusion coefficient and fractional anisotropy values in the developing infant brain. Am J Roentgenol 2010;195:W456–62. Provenzale JM, Liang L, Delong D, et al. Diffusion tensor imaging assessment of brain white matter maturation during the first postnatal year. Am J Roentgenol 2007;189:476–86. Chin EM, Choudhri AF. Myelination age: establishing normal ranges for myelination status during development using a histogram-based fractional anisotropy metric. In: Proceedings of the 51st Annual Meeting of the American Society of Neuroradiology, n.d. 2013. p. 280–81. Naidich TP, Duvernoy HM, Delman BN, et al. Duvernoy’s atlas of the human brain stem and cerebellum. New York: Springer; 2009. Nieuwenhuys R, Voogd J, Christiaan VH. The human central nervous system: a synopsis and atlas. New York: Springer; 2007. Mori S, van Zijl PC, Oishi K, et al. MRI atlas of human white matter. Waltham (MA): Academic Press; 2010. Schmahmann J, Pandya D. Fiber pathways of the brain. New York: Oxford University Press; 2009. Moreno-Jime´nez S, San-Juan D, La´rraga-Gutie´rrez JM, et al. Diffusion tensor imaging in radiosurgical callosotomy. Seizure 2012;21:473–7. Pizzini FB, Polonara G, Mascioli G, et al. Diffusion tensor tracking of callosal fibers several years after callosotomy. Brain Res 2010;1312:10–7.

23. Choudhri AF, Whitehead MT, McGregor AE, et al. Diffusion tensor imaging to evaluate commissural disconnection after corpus callosotomy. Neuroradiology 2013;55(11):1397–403. 24. Beaulieu C, Does MD, Snyder RE, et al. Changes in water diffusion due to wallerian degeneration in peripheral nerve. Magn Reson Med 1996;36:627–31. 25. Makki MI, Chugani DC, Janisse J, et al. Characteristics of abnormal diffusivity in normal-appearing white matter investigated with diffusion tensor MR imaging in tuberous sclerosis complex. AJNR Am J Neuroradiol 2007;28:1662–7. 26. Tillema JM, Leach JL, Krueger DA, et al. Everolimus alters white matter diffusion in tuberous sclerosis complex. Neurology 2012;78:526–31. 27. Rezaie R, Zouridakis G, Choudhri AF. Functional connectivity in tuberous sclerosis complex with autistic spectrum disorder preliminary findings. Journal of Pediatric Neurology 2013;11(2):79–82. 28. Simao G, Raybaud C, Chuang S, et al. Diffusion tensor imaging of commissural and projection white matter in tuberous sclerosis complex and correlation with tuber load. AJNR Am J Neuroradiol 2010;31:1273–7. 29. Shrivastava A. The hot cross bun sign. Radiology 2007;245:606–7. 30. Loh KB, Rahmat K, Lim SY, et al. A hot cross bun sign from diffusion tensor imaging and tractography perspective. Neurol India 2011;59:266–9. 31. Ito M, Watanabe H, Kawai Y, et al. Usefulness of combined fractional anisotropy and apparent diffusion coefficient values for detection of involvement in multiple system atrophy. J Neurol Neurosurg Psychiatry 2006;78:722–8. 32. De Simone T, Regna-Gladin C, Carriero M, et al. Wallerian Degeneration of the Pontocerebellar Fibers. AJNR Am J Neuroradiol 2005;26:1062–5.

425